External Email - Use Caution
Hi FS community,
I started using FS 7.1.0 for cortical and subcortical parcellation/segmentation.
As I found information that FLAIR data could improve pial surfaces creation, I decided to try it on since I have FLAIR images besides MPRAGE, by using the command line form FSwiki:
RECON-ALL -SUBJECT NAME -I INPUT_VOLUME.NII -FLAIR FLAIR_VOLUME.NII -FLAIRPIAL -ALL
However when I opened Freeview, I overlaid RH.PIAL, and RH.PIAL.FLAIR, RH.PIAL.T1.
The first two are quite similar (overlap) and the segmentation is worse than rh.pial.T1.
So, my question is, what is really the final surface of recon-all that must be checked for errors and that is used for stats estimation?
Thank you in advance.
Best regards,
Otília
rh.pial.T1 is the T1-based surface, rh.pial.FLAIR is the FLAIR-based surface refined from the T1 surface. rh.pial is the final surface and is a symbolic link to rh.pial.FLAIR
On 9/10/2020 10:40 AM, Otília C. d'Almeida wrote:
External Email - Use Caution
Hi FS community,
I started using FS 7.1.0 for cortical and subcortical parcellation/segmentation.
As I found information that FLAIR data could improve pial surfaces creation, I decided to try it on since I have FLAIR images besides MPRAGE, by using the command line form FSwiki:
*recon-all -subject NAME -i input_volume.nii -FLAIR FLAIR_volume.nii -FLAIRpial -all*
However when I opened Freeview, I overlaid *rh.pial*, and *rh.pial.FLAIR, rh.pial.T1*.
The first two are quite similar (overlap) and the segmentation is worse than rh.pial.T1.
So, my question is, what is really the final surface of recon-all that must be checked for errors and that is used for stats estimation?
Thank you in advance.
Best regards,
Otília
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
External Email - Use Caution
Thank you Douglas.
So, as far as I understood, the estimates (cortical thickness for example) are estimated from rh.pial surface (that comes from FLAIR pial surface). Since the use of FLAIR decreased the quality of pial segmentation (GM is underestimated since FLAIR pial surface was drawn throughout the GM tissue and not along the interface), should I re-run recon-all without -FLAIRpial flag or is there a way to use T1pial instead in order to obtain my estimates (I would prefer to check for small pial errors in this surface) without executing all recon-all steps? If it is possible, how?
Thanks,
Otília
A 2020-09-10 17:38, Douglas N. Greve escreveu:
rh.pial.T1 is the T1-based surface, rh.pial.FLAIR is the FLAIR-based surface refined from the T1 surface. rh.pial is the final surface and is a symbolic link to rh.pial.FLAIR
On 9/10/2020 10:40 AM, Otília C. d'Almeida wrote:
External Email - Use Caution
Hi FS community,
I started using FS 7.1.0 for cortical and subcortical parcellation/segmentation.
As I found information that FLAIR data could improve pial surfaces creation, I decided to try it on since I have FLAIR images besides MPRAGE, by using the command line form FSwiki:
RECON-ALL -SUBJECT NAME -I INPUT_VOLUME.NII -FLAIR FLAIR_VOLUME.NII -FLAIRPIAL -ALL
However when I opened Freeview, I overlaid RH.PIAL, and RH.PIAL.FLAIR, RH.PIAL.T1.
The first two are quite similar (overlap) and the segmentation is worse than rh.pial.T1.
So, my question is, what is really the final surface of recon-all that must be checked for errors and that is used for stats estimation?
Thank you in advance.
Best regards,
Otília
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
I think if you delete the ?h.pial files then run recon-all -s subject -autorecon3 -nosphere -nosurfreg -nowhite -nocortparc It should regenerate everything with the T1-only pial
On 9/10/2020 12:50 PM, Otília C. d'Almeida wrote:
External Email - Use Caution
Thank you Douglas.
So, as far as I understood, the estimates (cortical thickness for example) are estimated from rh.pial surface (that comes from FLAIR pial surface). Since the use of FLAIR decreased the quality of pial segmentation (GM is underestimated since FLAIR pial surface was drawn throughout the GM tissue and not along the interface), should I re-run recon-all without -FLAIRpial flag or is there a way to use T1pial instead in order to obtain my estimates (I would prefer to check for small pial errors in this surface) without executing all recon-all steps? If it is possible, how?
Thanks,
Otília
A 2020-09-10 17:38, Douglas N. Greve escreveu:
rh.pial.T1 is the T1-based surface, rh.pial.FLAIR is the FLAIR-based surface refined from the T1 surface. rh.pial is the final surface and is a symbolic link to rh.pial.FLAIR
On 9/10/2020 10:40 AM, Otília C. d'Almeida wrote:
External Email - Use Caution
Hi FS community,
I started using FS 7.1.0 for cortical and subcortical parcellation/segmentation.
As I found information that FLAIR data could improve pial surfaces creation, I decided to try it on since I have FLAIR images besides MPRAGE, by using the command line form FSwiki:
*recon-all -subject NAME -i input_volume.nii -FLAIR FLAIR_volume.nii -FLAIRpial -all*
However when I opened Freeview, I overlaid *rh.pial*, and *rh.pial.FLAIR, rh.pial.T1*.
The first two are quite similar (overlap) and the segmentation is worse than rh.pial.T1.
So, my question is, what is really the final surface of recon-all that must be checked for errors and that is used for stats estimation?
Thank you in advance.
Best regards,
Otília
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
Freesurfer mailing list Freesurfer@nmr.mgh.harvard.edu mailto:Freesurfer@nmr.mgh.harvard.edu https://mail.nmr.mgh.harvard.edu/mailman/listinfo/freesurfer
freesurfer@nmr.mgh.harvard.edu
-
Douglas N. Greve -
Otília C. d'Almeida